Nixtla/utilsforecast

Issues

• processing.counts_by_id returns a zero count for NaNs

  #91 opened 3 months ago by zecevicp
  5

• RMAE fails when trying to evaluate many models against one

  #101 opened 3 months ago by concimuscb
  1

• Calling RMAE on a crossvalidation DataFrame with some NaN forecasts RMAE evaluates them to 0.0

  #102 opened 3 months ago by concimuscb
  3

• Evaluate() to return aggregated metric evaluation across unique_ids.

  #87 opened 4 months ago by tackes
  0

• Using evaluate with MASE or RMAE

  #100 opened 4 months ago by concimuscb
  1

• Impossible to plot Insample Anomalies without a `forecast_df`

  #97 opened 4 months ago by omriKramer
  4

• MAPE calculation might need an update

  #86 opened 4 months ago by iamyihwa
  1

• Add Metric - Forecast Bias

  #88 opened 5 months ago by tackes
  0

• support distributed dataframes in evaluate

  #72 opened 7 months ago by AFilipaVEL
  0

• plotting.py raises ImportError for matplotlib, eventhough I'm using `plotly` engine

  #71 opened 7 months ago by emmandecena
  1

• plotting.py : _filter_series doesn't handle model names that break regex naming convention (eg. names that contain round parenthesis)

  #67 opened 8 months ago by paksas
  2

• Some evaluation metrics used the dataframe default index

  #62 opened 8 months ago by Jonathan-87
  7

• Polars deprecation warning `is_numeric`

  #55 opened 9 months ago by BCxP-Tobias
  1

• Smape shows only half of its value if Mord than one unique_ID is available.

  #49 opened a year ago by mkrech
  2

• [BUG] fill_gaps fails with quarterly frequency

  #48 opened a year ago by AzulGarza
  3

• [FEAT] support for pandas frequencies in generate_series

  #33 opened a year ago by AzulGarza
  1

• Overall Performance Metric

  #32 opened a year ago by baniasbaabe
  3

• Getting an error when using plot_series

  #31 opened a year ago by iamyihwa
  2